|
Activity Number:
|
573
- Simulation and Stochastic Bayesian Modeling
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #304940
|
Presentation
|
|
Title:
|
Bayesian Multi-Dimensional Functional Data Analysis
|
|
Author(s):
|
John Shamshoian* and Donatello Telesca
|
|
Companies:
|
UCLA School of Public Health and UCLA
|
|
Keywords:
|
Bayesian;
Functional Data Analysis;
Multi-Dimensional;
Weakly Separable;
Factor Analysis;
Gaussian Process
|
|
Abstract:
|
Multi-dimensional functional data arises in numerous modern scientific experiments or observational data. In this paper we focus on longitudinal functional data, a structured form of multidimensional functional data. Operating within a longitudinal functional framework is necessary to capture low dimensional interpretable features. We propose a computationally efficient nonparametric Bayesian method to simultaneously smooth observed data, estimate conditional functional means and functional covariance surfaces. Full distributions for quantities of interest can be simulated via an adaptive blocked Gibbs sampler. This method can also accommodate uncommon longitudinal points and uncommon functional grids. We compare our method to an existing method for longitudinal functional data in various simulation settings designed to mimic longitudinal functional data. We also apply our method to two case studies. The first case study involves age-specific fertility collected over time for various countries. The second case study is an implicit learning experiment in children with Autism Spectrum Disorder (ASD).
|
Authors who are presenting talks have a * after their name.
